Vaiyākaraṇa-siddhānta-kaumudī — bhāvakarmatiṅprakaraṇam

Learner edition · Prakaraṇa 63

63. Bhāvakarma-tiṅ-prakaraṇa — Verbs in bhāva and karma: yak, ciṇ and the ciṇvat-treatment

So far every verb you have built spoke of its agent: bhavati "he is", pacati "he cooks". A lakāra can also speak of the object (karman), or of the bare action (bhāva). This chapter builds those forms. After it you will be able to form bhūyate "there is becoming", with the agent in the third case: tvayā bhūyate. You will form anubhūyate ānandaḥ "bliss is experienced", where the ending speaks of the object. You will know the affix yak, which comes before a sārvadhātuka ending in these senses. You will know ciṇ, which replaces cli in the aorist: abhāvi. You will know the optional "treatment as before ciṇ" (ciṇvadbhāva), which gives doublets such as bhāvitā / bhavitā. Then come the root-specific adjustments: tāyate, anvatapta, dāyitā, ghānitā, grāhitā, śāmitā, aśami, abhāji, alambhi. The chapter ends with Bhaṭṭoji's verse on roots with two objects: which object the ending speaks of. Bhaṭṭoji calls the section the bhāvakarma-prakriyā, "the working of [verbs in] bhāva and karma". It has ten sūtras, SK 2756 to SK 2765.

Before you start

From chapter 1 you know vṛddhi (ā, ai, au) and guṇa (a, e, o). You know anuvṛtti, carrying a word forward from another sūtra, and upadeśa, the form as first taught. You know asiddha: treated as not having taken effect. You know the accents udātta and anudātta.
From chapter 3 you know sthānivat, "like the original": by SK 50 a substitute for a vowel may count as the vowel it replaced. SK 51 (na padānta…) switches this off in some operations, among them lengthening.
From chapter 8 you know the heading SK 200, aṅgasya: the rules under it change the aṅga, the base to which an affix is added. You know SK 254, aco ñṇiti: a vowel-final aṅga takes vṛddhi before an affix marked with ñ or ṇ.
From chapter 11 you know SK 354 (hali ca, lengthening before a consonant), SK 358 (the h of han becomes gh) and SK 415 (loss of a root's n before a k- or ṅ-marked affix).
From chapter 16 you know the participants (kāraka): the agent (kartṛ) and the object (karman). You know that an agent not expressed by the verb takes the third case. You know SK 539–541, which give the name karman to a second object.
From chapter 44 you know the verb machinery. SK 2152 lets a lakāra denote the karma (after a root that takes an object) or bhāva (after one that does not). You know the persons (SK 2165), the names sārvadhātuka (SK 2166) and ārdhadhātuka (SK 2187), and the augment iṭ (SK 2184). You know cli and sic (SK 2222), and ciṇ with deletion of the following ta (SK 2328–2329).
From chapter 44 you also know these: sya (SK 2186), the luṭ affix tās, sīyuṭ (SK 2255), eś for ta in liṭ (SK 2241) and ha eti (SK 2250). You know SK 2282 (vṛddhi of a penultimate a), SK 2298, SK 2300, SK 2305 and SK 2313 (ṇer aniṭi). You know SK 2319, 2364, 2366, 2368, 2370, 2380, 2389–2391, 2405, 2407 and 2409.
From chapter 44 you also know SK 2233 (te for ta), SK 2247 (ḍh for dh), SK 2262 (the augment num), SK 2281 (jhalo jhali, loss of an s between two consonants) and SK 2372 (āto lopa iṭi ca, loss of ā).
From chapter 45 you know SK 2433 (vadha for han before liṅ), SK 2462, SK 2473 and SK 2480. From chapter 47 you know SK 2512 and SK 2513 (ciṇ te padaḥ). From chapter 48 you know SK 2528. From chapter 50 you know SK 2544. From chapter 51 you know SK 2550: after sam, kṛ takes an s before it (saṃskṛ). From chapter 52 you know SK 2562 (graho 'liṭi dīrghaḥ).
From chapter 53 you know the mit roots, whose penultimate vowel stays short before ṇi (SK 2568). You know SK 2574, hanas to 'ciṇṇaloḥ. From chapter 54 you know the causative ṇic, SK 2582 (labheś ca) and SK 2604.
From chapters 55 to 57 you know the desiderative (san), the intensive (yaṅ, SK 2629) and the intensive whose yaṅ is lost (yaṅluk). You know SK 2649, ayaṅ yi kṅiti.
From chapter 61 you know SK 2679, bhāvakarmaṇoḥ: a lakāra in bhāva or karma takes the ātmanepada endings (taṅ). Every form in this chapter is ātmanepada for that reason.

A. yak in bhāva and karma: the impersonal third singular — running example: `bhū + laṭ (bhāva) → ta (taṅ, SK 2679) → bhū + yak + te → bhūyate`; `tvayā mayā 'nyaiś ca bhūyate`

Bhaṭṭoji opens the section: "Now laṭ and the other lakāras in bhāva and karma. By bhāvakarmaṇoḥ (SK 2679), the taṅ endings come." One sūtra follows. It puts the affix yak between the root and a sārvadhātuka ending. The running example is bhū "to be, to become". It takes no object, so its lakāra may denote bhāva, the bare action. "There is becoming by you, by me and by others" is tvayā mayā 'nyaiś ca bhūyate.

SK 2756 · सार्वधातुके यक् · sārvadhātuke yak (A 3.1.67)

सार्वधातुके यक्

sārvadhātuke yak · A 3.1.67

धातोर्यक् प्रत्ययः स्याद्भावकर्मवाचिनि सार्वधातुके परे । भावो भावना, उत्पादना क्रिया सा च धातुत्वेन सकलधातुवाच्या भावार्थकलकारेणानूद्यते । युष्मदस्मद्भ्यां समानाधिकरण्याभावात्प्रथमपुरुषः । तिङ्वाच्यभावनाया असत्त्वरूपत्वेन द्वित्वाद्यप्रतीतेर्न द्विवचनादि । किंत्वेकवचनमेव । तस्यौत्सर्गिकत्वेन संख्यानपेक्षत्वात् ॥ अनभिहिते कर्तरि तृतीया । त्वया मयाऽन्यैश्च भूयते । बभूवे ॥

Rule. The affix yak comes after a root when a sārvadhātuka affix follows that denotes bhāva or karma. Only ya is heard; the k is a marker. Bhaṭṭoji's gloss: "let the affix yak come after a dhātu when there follows a sārvadhātuka denoting bhāva or karma".

"After a dhātu" (dhātoḥ) is carried forward from SK 2629. "Denoting bhāva or karma" is carried forward from SK 2758, two sūtras later. Bhaṭṭoji's commentators both give these.

Example.

bhū + laṭ, in bhāva (SK 2152; bhū takes no object)

→ the ātmanepada by SK 2679; third person singular → bhū + ta

→ the ta becomes te (SK 2233) → bhū + te

→ te is a sārvadhātuka denoting bhāva, so this rule adds yak → bhū + ya + te

→ yak is marked with k, and such an affix blocks guṇa (chapter 44) → bhūyate.

The agent is not expressed by the ending (anabhihita), so it takes the third case: tvayā mayā 'nyaiś ca bhūyate.

In liṭ there is no sārvadhātuka, so there is no yak: babhūve.

Applies when / blocked by. What is bhāva? Bhaṭṭoji: bhāva is bhāvanā, "bringing about" (utpādanā), that is, action (kriyā). Every root already denotes action, just by being a root. So a lakāra in the sense of bhāva only restates (anūdyate) what the root has said. One of Bhaṭṭoji's commentators compares the dual ending in dvau "two", which restates the twoness the word already names.

Why only the third person? Bhaṭṭoji: there is no co-reference (sāmānādhikaraṇya) with yuṣmad "you" or asmad "I". One of his commentators explains. A lakāra can share its referent with a word only when it denotes the agent or the karma, never bhāva. The third person still comes by SK 2165, śeṣe prathamaḥ: it is the person used wherever the second and first do not apply.

Why only the singular? Bhaṭṭoji: the action denoted by a verb-ending is not a substance (asattva). So two or many of it are not understood. The singular is used because it is the general ending (autsargika). It does not depend on the number one. One of Bhaṭṭoji's commentators contrasts the action denoted by a kṛt affix, which is like a substance: pākau, pākāḥ "two cookings, many cookings". Whether "only the singular" holds without exception, scholars discuss this.

Watch out. Yak comes only before a sārvadhātuka: the endings of laṭ, loṭ, laṅ and vidhiliṅ. The other lakāras take no yak. Liṭ is formed as usual: babhūve. Before sya, sic, sīyuṭ and tās, SK 2757 optionally applies. In the luṅ before ta, SK 2758 brings ciṇ.

B. ciṇ in the luṅ and the ciṇvat-treatment, with a survey of yak forms — running example: `bhū + luṅ (bhāva), ta → cli → ciṇ → (vṛddhi, ta luk) → abhāvi`; `bhū + tās (ciṇvat: vṛddhi + iṭ) → bhāvitā / bhavitā`

Two sūtras make the forms that have no yak. SK 2758 replaces cli by ciṇ in the aorist before ta: abhāvi. SK 2757 comes first in Bhaṭṭoji's order. It lets the aṅga be treated, optionally, as if ciṇ followed it. This happens before sya (the affix of lṛṭ and lṛṅ), sic (the aorist), sīyuṭ (the liṅ of blessing) and tās (the periphrastic future). With it comes an augment iṭ. So bhū in luṭ gives bhāvitā, with vṛddhi as before ciṇ, or bhavitā, as usual. Under SK 2758 Bhaṭṭoji then surveys yak forms: the karma, the causative, the desiderative, the intensive, and roots whose stem changes before yak.

SK 2757 · स्यसिच्सीयुट्तासिषु भावकर्मणोरुपदेशेऽज्झनग्रहदृशां वा चिण्वदिट् च · syasicsīyuṭtāsiṣu bhāvakarmaṇorupadeśe'jjhanagrahadṛśāṃ vā ciṇvadiṭ ca (A 6.4.62)

स्यसिच्सीयुट्तासिषु भावकर्मणोरुपदेशेऽज्झनग्रहदृशां वा चिण्वदिट् च

syasicsīyuṭtāsiṣu bhāvakarmaṇorupadeśe'jjhanagrahadṛśāṃ vā ciṇvadiṭ ca · A 6.4.62

उपदेशे योऽच् तदन्तानां हनादीनां च चिणीवाङ्गकार्यं वा स्यात्स्यादिषु परेषु भावकर्मणोर्गम्यमानयोः स्यादीनामिडागमश्च । अयमिट् चिण्वद्भावसन्नियोगशिष्टत्वात्तदभावे न । इहार्धधातुक इत्यधिकृतं सीयुटो विशेषणं नेतरेषामव्यभिचारात् । चिण्वद्भावाद्वृद्धिः । भाविता । भविता । भाविष्यते । भविष्यते । भूयताम् । अभूयत । भूयेत । भाविषीष्ट । भविषीष्ट ॥

Rule. Take an aṅga that ends in a vowel (ac) which stood as such in upadeśa, or the aṅga of han, grah or dṛś. When sya, sic, sīyuṭ or tās follows, and bhāva or karma is understood, that aṅga is optionally treated as before ciṇ. And sya and the rest take the augment iṭ.

Bhaṭṭoji's gloss: "let there optionally be the operation on the aṅga (aṅgakārya) that takes place before ciṇ … and the augment iṭ for sya and the rest". The word ciṇvat means "as before ciṇ" (ciṇīva). "Of the aṅga" comes from the heading SK 200. Bhaṭṭoji's commentators both give these points.

Example.

bhū + luṭ, in bhāva → bhū + tās + …

→ this rule, first option: the aṅga is treated as before ciṇ. Ciṇ is marked with ṇ, so the final ū takes vṛddhi (SK 254) → bhau

→ tās takes the joint iṭ; au before a vowel becomes āv → bhāv-i-tā

→ bhāvitā.

Second option, without the treatment:

bhū + tās → the ordinary iṭ before a val-initial affix (SK 2184) → guṇa → bhavitā.

Likewise lṛṭ bhāviṣyate / bhaviṣyate, and the liṅ of blessing bhāviṣīṣṭa / bhaviṣīṣṭa.

Before a sārvadhātuka the rule has no place, and yak comes instead: loṭ bhūyatām, laṅ abhūyata, vidhiliṅ bhūyeta.

Applies when / blocked by. Bhaṭṭoji makes three points.

First, the iṭ is enjoined jointly (sanniyogaśiṣṭa) with the treatment. Where the treatment is absent, this iṭ is absent. One of his commentators adds that a seṭ root still takes its ordinary iṭ then. That is why the second option is bhavitā.

Second, the word "before an ārdhadhātuka" is in force here as a heading. It qualifies sīyuṭ alone. Sya, sic and tās are always ārdhadhātuka, so it adds nothing for them. For sīyuṭ it keeps the rule out of the vidhiliṅ: bhūyeta. Bhaṭṭoji's commentators both draw this out.

Third, the treatment brings vṛddhi: bhāvitā.

Bhaṭṭoji's gloss reads upadeśe yo 'c, "a vowel that stands in upadeśa". Both of his commentators explain that "in upadeśa" qualifies the vowel only, not han and the rest. The Tattvabodhinī adds the contrast: the gloss is not "a root that ends in a vowel in upadeśa".

One of them, the Bālamanoramā, adds two points. "When bhāva or karma is understood" is a seventh case of domain (viṣayasaptamī), since sya, sic and tās themselves denote neither. And the four affixes are not paired one by one with the kinds of root.

What does the treatment bring? Both commentators cite a verse that lists its uses. Its opening words name vṛddhi, the augment yuk, and gh for the h of han. You will meet all three in this chapter. Both also say that only operations on the aṅga are brought. So vadha for han is not brought (see SK 2761).

When the treatment is taken, its iṭ blocks the ordinary iṭ. Bhaṭṭoji's commentators both say why: it is nitya, a rule that applies whatever is done first.

Watch out. The rule is optional, and the iṭ belongs to sya, sic, sīyuṭ and tās, not to the root. Always give both forms: bhāvitā / bhavitā.

SK 2758 · चिण् भावकर्मणोः · ciṇ bhāvakarmaṇoḥ (A 3.1.66)

चिण् भावकर्मणोः

ciṇ bhāvakarmaṇoḥ · A 3.1.66

च्लेश्चिण् स्वाद्भावकर्मवाचिनि तशब्दे परे । अभावि । अभाविष्यत । अभविष्यत । तिङोक्तत्वात्कर्मणि न द्वितीया । अनुभूयते आनन्दश्चैत्रेण त्वया मया च । अनुभूयेते । अनुभूयन्ते । त्वमनुभूयसे । अहमनुभूये । अन्वभावि । अन्वभाविषाताम् । अन्वभविषाताम् । णिलोपः । भाव्यते । भावयांचक्रे । भावयांबभूवे । भावयामासे । इह तशब्दस्य एशि इट एत्वे च कृते ह एतीति हत्वं न । तासिसाहचर्यादस्तेरपि व्यतिहे इत्यादौ सार्वधातुके एव ह एति SK 2250 इति हत्वप्रवृत्तेरित्याहुः । भाविता । चिण्वदिट आभीयत्वेनासिद्धत्वाण्णिलोपः । पक्षे भावयिता । भाविष्यते । भावयिष्यते । भाव्यताम् । अभाव्यत । भाव्येत । भाविषीष्ट । भावयिषीष्ट । अभावि । अभाविषाताम् । अभावयिषाताम् । बुभूप्यते । बुभूषांचक्रे । बुभूषिता । बुभूषिष्यते । बोभूय्यते । यङ्लुगन्तात्तु । बोभूयते । बोभवांचक्रे । बोभाविता । बोभविता । अकृत्सार्व-SK 2298 इति दीर्घः । स्तूयते विष्णुः । तुष्टुवे । स्ताविता । स्तोता । स्ताविष्यते । स्तोष्यते । अस्तावि । अस्ताविषाताम् । अस्तोषाताम् । गुणोर्ति SK 2380 इति गुणः । अर्यते । स्मर्यते । सस्मरे । परत्वान्नित्यत्वाच्च गुणे रपरे कृतेऽजन्तत्वाभावेऽप्युपदेशग्रहणाच्चिण्वदिट् । आरिता । अर्ता । स्मारिता । स्मर्ता । गुणोऽर्ति-SK 2380 इत्यत्र नित्यग्रहणानुवृत्तेरुक्तत्वान्नेह गुणः । संस्क्रियते । अनिदिताम्-SK 415 इति नलोपः । स्रस्यते । इदितस्तु नन्द्यते । संप्रसारणम् । इज्यते । अयङ् यिक्ङिति SK 2649 । शय्यते ॥

Rule. Ciṇ replaces cli when the ending ta follows, denoting bhāva or karma. Only i is heard. "In place of cli" is carried forward from SK 2222, and "before ta" from SK 2513. Bhaṭṭoji's commentators both give these.

Example.

bhū + luṅ, in bhāva → the augment aṭ, cli, the ending ta → a + bhū + cli + ta

→ this rule: ciṇ replaces cli → a + bhū + i + ta

→ ciṇ is marked with ṇ: vṛddhi of ū (SK 254), and āv before the vowel → a + bhāv + i + ta

→ the ta after ciṇ is deleted (SK 2329) → abhāvi.

Lṛṅ has sya, so SK 2757 gives two forms: abhāviṣyata / abhaviṣyata.

Applies when / blocked by. Only before ta. Before the dual ātām there is no ciṇ. Sic stays, with the option of SK 2757: anvabhāviṣātām / anvabhaviṣātām. One of Bhaṭṭoji's commentators gives this as the reason for "before ta".

The karma. Bhaṭṭoji now turns from bhāva to the karma. Anu + bhū "to experience" takes an object. In a karma lakāra the ending expresses the object. So the object does not take the second case:

anubhūyate ānandaś caitreṇa tvayā mayā ca "bliss is experienced by Caitra, by you and by me".

Other persons and numbers are found too:

anubhūyete (supply sukhaduḥkhe, "pleasure and pain"), anubhūyante (supply sukhāni, "pleasures");
tvam anubhūyase "you are experienced", aham anubhūye "I am experienced";
luṅ anvabhāvi; dual anvabhāviṣātām / anvabhaviṣātām.

One of Bhaṭṭoji's commentators supplies these subjects.

The causative. From bhāvi (bhū + ṇic):

bhāvi + yak + te → the ṇi is deleted before yak (SK 2313) → bhāvyate.

Liṭ: bhāvayāṃcakre, bhāvayāṃbabhūve, bhāvayāmāse.

In bhāvayāmāse, eś replaces ta (SK 2241). The same form is also the first person, where e replaces the ending iṭ (SK 2233). Should the s of ās then become h before e by SK 2250? No, Bhaṭṭoji says, both for eś and for the e that replaces iṭ. SK 2250 gives h to tās only before a sārvadhātuka e. By association with tās, the same holds for as. Bhaṭṭoji gives this as others' view: "so they say" (ity āhuḥ). One of his commentators adds that this e is ārdhadhātuka.

Luṭ of the causative:

bhāvi + tās, treated as before ciṇ → the joint iṭ comes → bhāvi + i + tā

→ SK 2313 deletes ṇi only before an ārdhadhātuka without iṭ. But this iṭ belongs to the ābhīya section of rules, and counts there as asiddha. So for SK 2313 it is not there

→ the ṇi is deleted → bhāvitā.

In the other option, the ordinary iṭ is not asiddha. So the ṇi stays, and takes guṇa → bhāvayitā.

Bhaṭṭoji's commentators both draw a lesson. This is why SK 2757 is read as "a vowel in upadeśa" and not "a root ending in a vowel in upadeśa". The stem bhāvi was never taught as such, yet it takes the treatment.

Likewise: bhāviṣyate / bhāvayiṣyate; bhāvyatām, abhāvyata, bhāvyeta; bhāviṣīṣṭa / bhāvayiṣīṣṭa; abhāvi; abhāviṣātām / abhāvayiṣātām.

The desiderative. bubhūṣyate, bubhūṣāṃcakre, bubhūṣitā, bubhūṣiṣyate.

The intensive. With yaṅ: bobhūyyate. With yaṅ lost: bobhūyate, bobhavāṃcakre, bobhāvitā / bobhavitā.

Stems that change before yak. Bhaṭṭoji ends with five kinds.

Lengthening by SK 2298, which lengthens a final vowel before a y-initial affix that is neither kṛt nor sārvadhātuka: stu + ya + te → stūyate viṣṇuḥ "Viṣṇu is praised". Also tuṣṭuve; stāvitā / stotā; stāviṣyate / stoṣyate; astāvi; astāviṣātām / astoṣātām.
Guṇa by SK 2380, guṇo 'rtisaṃyogādyoḥ: ṛ + ya + te → aryate; smṛ → smaryate. Bhaṭṭoji also gives the liṭ sasmare.
No guṇa in saṃskriyate. Bhaṭṭoji's reason: it has been stated that "invariably" (nitya) is carried into SK 2380. One of his commentators explains. SK 2380 then acts only on a root that always begins with a cluster, and kṛ begins with one only when s is added after sam (SK 2550).
Loss of n by SK 415: sraṃs → srasyate. But a root marked with i (idit) keeps its n: nandyate.
Saṃprasāraṇa: yaj → ijyate. One commentator names SK 2409 for it.
SK 2649, ayaṅ yi kṅiti: śī → śayyate.

Applies when / blocked by. The ṛ-roots raise a question. Before tās, the ṛ first takes guṇa, with r added (rapara). Bhaṭṭoji says why: it is the later rule (para), and it is nitya. But then the base ends in r, not in a vowel. Does SK 2757 still apply? Yes, Bhaṭṭoji says, "because upadeśa is mentioned": the vowel stood in upadeśa. So āritā / artā, smāritā / smartā. Whether smāritā is fully saved by this gloss, scholars discuss this.

Watch out. In the karma the ending expresses the object, so other persons and numbers are found: anubhūyete, tvam anubhūyase, aham anubhūye. In bhāva only the third person singular is used (SK 2756).

C. Two root-specific adjustments, to yak and to ciṇ — running example: `tan + yak → (ā for n, optional) → tāyate / tanyate`; `anu + tap, luṅ → (no ciṇ) sic → anvatapta pāpena`

Two sūtras adjust single roots. SK 2759 lets tan "to stretch" replace its final n by ā before yak: tāyate, or tanyate. SK 2760 forbids ciṇ to tap "to burn" in two settings: when the karma is itself the agent, and in the sense of remorse. So "he was struck by sin" is anvatapta pāpena, with sic, not ciṇ. Under SK 2760 Bhaṭṭoji adds two more stem changes before yak, and one in liṭ.

SK 2759 · तनोतेर्यकि · tanoteryaki (A 6.4.44)

तनोतेर्यकि

tanoteryaki · A 6.4.44

आकारोऽन्तादेशो वा स्यात् । तायते । तन्यते । ये विभाषा SK 2391 । जायते । जन्यते ॥

Rule. When yak follows, ā optionally replaces the final sound of tan. "Ā" is carried forward from SK 2982, taught later, and "optionally" (vibhāṣā) from ye vibhāṣā (SK 2319). Bhaṭṭoji's commentators both give these. Bhaṭṭoji's gloss fills in the rest: "let ā optionally be the substitute for the final".

Example.

tan + yak + te (in the karma: "it is stretched")

→ this rule: ā replaces the final n → tā + ya + te

→ tāyate.

In the other option the n stays → tanyate.

Bhaṭṭoji adds the parallel case by SK 2319 itself: jāyate / janyate.

Applies when / blocked by. Why "before yak"? One of Bhaṭṭoji's commentators gives the intensive taṃtanyate. There the y is that of yaṅ, so the rule does not act.

Watch out. Only yak triggers the rule. A ya of another affix leaves tan alone.

SK 2760 · तपोऽनुतापे च · tapo'nutāpe ca (A 3.1.65)

तपोऽनुतापे च

tapo'nutāpe ca · A 3.1.65

तपश्चलेश्चिण्न स्यात्कर्मकर्तर्यनुतापे च । अन्वतप्त पापेन । पापं कर्तृ । तेनाभ्याहत इत्यर्थः । कर्मणि लुङ् । यद्वा पापेन पुंसा कर्त्रा अशोचीत्यर्थः । घुमास्था-SK 2462 इतीत्वम् । दीयते । धीयते । आदेचः-SK 2370 इत्यत्राशितीति कर्मधारयादित्संज्ञकशकारादौ निषेधः । एष आदिशित्त्वाभावान्न तस्मिन् आत्वम् । जग्ले ॥

Rule. For tap, ciṇ does not replace cli when the karma is the agent (karmakartṛ). And it does not replace it in the sense of remorse (anutāpa). Bhaṭṭoji's gloss: "for tap, ciṇ does not replace cli, when the karma is agent, and in the sense of remorse".

The sūtra says only "for tap, in remorse, and". The rest is carried forward. "Ciṇ" comes from SK 2513 and "in place of cli" from SK 2222. "Not" (na) comes from na rudhaḥ (SK 2770), taught later. The word ca "and" draws in "when the karma is agent" from acaḥ karmakartari (SK 2768), also taught later. Bhaṭṭoji's commentators both give this, the Bālamanoramā in full.

Example.

anu + tap + luṅ, third person singular → anu + a + tap + cli + ta

→ SK 2758 would bring ciṇ; this rule forbids it in the sense of remorse

→ sic (SK 2222) stays as the replacement of cli → anu + a + tap + s + ta

→ the s of sic is lost between two consonants, p and t (SK 2281, jhalo jhali) → anvatapta

→ anvatapta pāpena.

Bhaṭṭoji gives two readings of the sentence.

Pāpa "sin" is the agent, and the meaning is "he was struck by it". The luṅ is in the karma: the man is the object struck. One of Bhaṭṭoji's commentators says why: with anu, tap here means "to strike", which takes an object.
Or pāpena means "by a sinful man", as agent, and the meaning is "there was grieving" (aśocīt). Then the luṅ is in bhāva. Bhaṭṭoji's commentators both take it so. One of them adds that grieving takes no object.

The example for "when the karma is agent" comes in chapter 64. One of Bhaṭṭoji's commentators says so.

Two forms before yak, and one in liṭ. Bhaṭṭoji adds these under this sūtra.

Ī for the ā of dā and dhā by SK 2462: dā + ya + te → dīyate; dhīyate.
Glai "to be weary", in liṭ. Eś replaces ta (SK 2241). Does SK 2370 give ā for ai here? It is barred "before a śit" (aśiti). Bhaṭṭoji reads aśiti as a karmadhāraya compound: "before an affix whose initial is an ś that is a marker". Eś has its ś at the end, not the beginning. So the bar does not hold, and ā comes:

glai + eś → glā + e → doubling → jaglā + e → the ā is lost before the vowel-initial ending → jagle.

One of Bhaṭṭoji's commentators gives this construal: no bar, so ā does come before eś. He names āto lopa iṭi ca (SK 2372) for the loss of ā.

Watch out. This rule only forbids. The sic that stays is the ordinary aorist affix of chapter 44.

D. yuk for ā-finals, and the ciṇvat forms of han, grah, dṛś and gṝ — running example: `dā + tās (ciṇvat) → dā-y-i-tā (yuk) → dāyitā / dātā`; `dā + luṅ ciṇ → adāyi`

One sūtra here, SK 2761, adds the augment yuk, heard as y, to an aṅga ending in ā. It acts before ciṇ, and so also wherever SK 2757 treats the aṅga as before ciṇ. The running example is dā "to give": dāyitā "it will be given", or dātā; aorist adāyi. Under this sūtra Bhaṭṭoji works through the other three roots named in SK 2757: han, grah and dṛś. He then counts the hundred and four aorist forms of gṝ "to swallow". He ends with a causative in the karma, śamyate, which leads into the next group.

SK 2761 · आतो युक् चिण्कृतोः · āto yuk ciṇkṛtoḥ (A 7.3.33)

आतो युक् चिण्कृतोः

āto yuk ciṇkṛtoḥ · A 7.3.33

आदन्तानां युगागमः स्याच्चिणि ञिति णिति कृति च । दायिता । दाता । दायिषीष्ट । दासीष्ट । अदायि । अदायिषाताम् । स्थाघ्वोरिच्च SK 2389 । अदिषाताम् । अधायिषाताम् । अधिषाताम् । अग्लायिषाताम् । अग्लासाताम् । इत्यते । अचिण्णलोः SK 2574 इत्युक्तेर्हनस्तो न । हो हन्तेः-SK 358 इति कुत्वम् । घानिता । हन्ता । धानिष्यते । हनिष्यते । आशीर्लिङि वधादेशस्यापवादश्चिण्वद्भावः । आर्धधातुके सीयुटीति विशेषविहितत्वात् । घानिषीष्ट । पक्षे वधिषीष्ट । अघानि । अघानिषाताम् । अहसाताम् । पक्षे वधादेशः । अवधि । अवधिषाताम् । अघानिष्यत । अहनिष्यत । न च स्यादिषु चिण्वदित्यतिदेशाद्वधादेशः स्यादिति वाच्यम् । अङ्गस्य SK 200 इत्यधिकारादाङ्गस्यैवातिदेशात् । गृह्यते । चिण्वदिटो न दीर्घत्वम् । प्रकृतस्य वलादिलक्षणस्यैवेटो ग्रहोऽलिटि-SK 2562 इत्यनेन दीर्घविधानात् । ग्राहिता । ग्रहीता । ग्राहिष्यते । ग्रहीष्यते । ग्राहिषीष्ट । ग्रहिषीष्ट । अग्राहि । अग्राहिषातम् । अग्रहीषाताम् । दृश्यते । अदर्शि । अदर्शिषाताम् । सिचः कित्त्वादम्न । अदृक्षाताम् । गिरतेर्लुङि ध्वमि चतुरधिकं शतम् । तथा हि चिण्वदिटो दीर्घो नेत्युक्तम् । अगारिध्वम् । द्वितीये त्विटि वॄतो वा SK 2391 इति वा दीर्घः । अगरीध्वम् । अगरिध्वम् । एषां त्रायाणां लत्वं ढत्वं द्वित्वत्रयं चेति पञ्च वैकल्पिकानि । इत्थं षण्णवतिः । लिङ्सिचोः SK 2528 इति विकल्पत्वादिडभावे उश्च SK 2368 इति कित्त्वम् । रपरत्वं हलि च SK 354 इति दीर्घः । इणः षीध्वं-SK 2247 इति नित्यं ढत्वम् । अशीढ्र्वम् । ढवमानां द्वित्वविकल्पे अष्टौ । उक्तषण्णवत्या सह संकलने उक्ता संख्येति ॥ इङ् दीर्घश्चिण्वदिट् लत्वं ढत्वं द्वित्वत्रिकं तथा । इत्याष्टानां विकल्पेन चतुर्भिरधिकं शतम् ॥ हेतुमण्ण्यन्तात्कर्मणि लः । यक् । णिलोपः । शभ्यते मोहो मुकुन्देन ॥

Rule. An aṅga ending in ā takes the augment yuk when ciṇ follows, or a kṛt affix marked with ñ or ṇ. The u and k are markers; y is added. "Marked with ñ or ṇ" (ñṇiti) is carried forward from SK 254.

Bhaṭṭoji's commentators both say that ñṇiti qualifies the kṛt affix only. Ciṇ is always marked with ṇ, so the word adds nothing for it.

Example.

dā + luṭ, in the karma → dā + tās + …

→ SK 2757: treated as before ciṇ, with the joint iṭ

→ this rule: yuk before ciṇ, here by that treatment → dā + y + i + tā

→ dāyitā.

Without the treatment there is no yuk, and dā takes no iṭ → dātā.

Likewise: dāyiṣīṣṭa / dāsīṣṭa. Luṅ with ciṇ itself: dā + ciṇ → yuk → adāyi.

Dual of the luṅ, with sic: adāyiṣātām. Without the treatment, SK 2389 (sthāghvor icca) gives i for the ā: adiṣātām. Bhaṭṭoji's commentators both explain that the joint iṭ, as the later rule, blocks SK 2389 in the first form.

So too dhā: adhāyiṣātām / adhiṣātām; and glai: aglāyiṣātām / aglāsātām.

Applies when / blocked by. One of Bhaṭṭoji's commentators gives counter-examples. Why "before ciṇ or a kṛt"? cauḍiḥ, bālākiḥ (a taddhita affix iñ) and dadau (a liṭ ending) take no yuk. Why "marked with ñ or ṇ"? pānīyam, dānīyam.

han. Present hanyate. With the treatment:

han + tās → treated as before ciṇ

→ SK 2574 replaces the n of han by t, but not before ciṇ. So here too the n stays

→ gh for h by SK 358, reached through the treatment

→ vṛddhi of the penultimate a, with the joint iṭ → ghānitā.

Without the treatment: hantā. Likewise ghāniṣyate / haniṣyate.

Liṅ of blessing. SK 2433 replaces han by vadha before liṅ. But here the treatment wins: it is the exception (apavāda) to vadha. Bhaṭṭoji's reason: it is specifically enjoined in the clause "before sīyuṭ, an ārdhadhātuka". So ghāniṣīṣṭa; in the other option, vadha: vadhiṣīṣṭa.

Bhaṭṭoji's commentators explain "specifically" in two ways.

The Bālamanoramā reads the rule as four clauses for each root: before sya, before sic, before sīyuṭ, before tās. For han the sīyuṭ-clause has no other place to act (niravakāśa). So it must set vadha aside.
The Tattvabodhinī compares the two triggers. Liṅ belongs to both sets of endings; sīyuṭ comes only in the ātmanepada. So the treatment is the more specific rule.

Luṅ: aghāni. Dual aghāniṣātām / ahasātām. In the other option, vadha: avadhi, avadhiṣātām. Lṛṅ: aghāniṣyata / ahaniṣyata.

Could the treatment itself bring vadha in sya and the rest? No, says Bhaṭṭoji. Because of the heading SK 200, only an operation on the aṅga is extended. One of his commentators adds that vadha is taught in the second book, outside that heading.

grah. Present gṛhyate. The joint iṭ is not lengthened. SK 2562 lengthens only the iṭ in its own context: the ordinary iṭ before a val-initial affix. So:

grāhitā / grahītā; grāhiṣyate / grahīṣyate; agrāhi; agrāhiṣātām / agrahīṣātām.

In the liṅ of blessing the text gives grāhiṣīṣṭa / grahiṣīṣṭa; whether the second should be long, as SK 2562 would make it, scholars discuss this.

dṛś. Present dṛśyate. Luṅ adarśi; dual adarśiṣātām. Without the treatment: sic is marked as kit by SK 2300. So the augment am of SK 2405 does not come, since it needs an affix that is not kit: adṛkṣātām.

gṝ in the luṅ, before dhvam: a hundred and four forms. Bhaṭṭoji counts them.

With an iṭ there are three bases. The joint iṭ, not lengthened, gives agāridhvam. The ordinary iṭ, optionally long by SK 2391, gives agarīdhvam and agaridhvam.
Each of the three takes five optional operations. They are l for r (latva), ḍh for dh (ḍhatva), and three doublings. Each option doubles the count: three bases × 2 × 2 × 2 × 2 × 2 make ninety-six.
The ordinary iṭ is itself optional here (SK 2528). Without it, sic is kit by SK 2368. The ṝ becomes i, with r after it. Bhaṭṭoji does not name the step ṝ → i; one of his commentators supplies it (SK 2390). The i is lengthened by SK 354. Ḍh comes, now invariably, by SK 2247. So agīrḍhvam.
That form allows three optional doublings, of ḍh, v and m: eight forms.
Ninety-six and eight make a hundred and four.

Bhaṭṭoji sums it up in a verse. The eight options are the iṭ, its lengthening, the joint iṭ, latva, ḍhatva and the three doublings. Both of his commentators set out the count step by step, in slightly different orders, and name the rule for each option.

Watch out. Keep the two *iṭ*s apart. With the joint iṭ of SK 2757, the root takes vṛddhi from the treatment, and the iṭ is not lengthened: grāhitā, agāridhvam. With the ordinary iṭ, the root takes guṇa. That iṭ is long for grah (SK 2562): grahītā. It is optionally long for gṝ (SK 2391): agarīdhvam / agaridhvam.

A causative in the karma. Bhaṭṭoji closes the sūtra with śam "to be calm" + the causative ṇic (his hetumaṇṇic). Its lakāra is in the karma; yak comes; the ṇi is deleted:

śami + yak + te → śamyate moho mukundena "delusion is calmed by Mukunda".

One of Bhaṭṭoji's commentators explains why the karma: a causative stem always takes an object.

E. The upadhā vowel before ciṇ: optional length for mit roots, no vṛddhi for m-final udātta roots — running example: `śam + ṇic + tās (ciṇvat) → śāmitā / śamitā / śamayitā`; `śam + luṅ ciṇ → (no vṛddhi) aśami`

Both sūtras here act on the upadhā, the penultimate vowel, before ciṇ. SK 2762 is for the causative of a mit root. Its a was kept short before ṇi by SK 2568. Before ṇi followed by ciṇ it may now be long: śāmitā or śamitā, beside śamayitā without the treatment. SK 2763 is for the plain root. An m-final root that is udātta in upadeśa takes no vṛddhi before ciṇ: aśami, not aśāmi. The running example is śam "to be calm" in both.

SK 2762 · चिण्णमुलोर्दीर्घोऽन्यतरस्याम् · ciṇṇamulordīrgho'nyatarasyām (A 6.4.93)

चिण्णमुलोर्दीर्घोऽन्यतरस्याम्

ciṇṇamulordīrgho'nyatarasyām · A 6.4.93

चिण् परे णमुल्परे च णौ मितामुपधाया दीर्घो वा स्यात् । प्रकृतो मितां ह्रस्वः-SK 2568 एव तु न विकल्पितः । ण्यन्ताण्णौ ह्रस्वविकल्पस्यासिद्धेः । दीर्घविधौ हि णिचो लोपो न स्थानिवदिति दीर्घः सिध्यति । ह्रस्वविधौ तु स्थानिवत्त्वं दुर्वारम् । भाष्ये तु पूर्वत्रासिद्धीये न स्थानिवदित्यवष्टभ्य द्विर्वचनसवर्णानुस्वारदीर्घजश्चरः प्रत्याख्याताः । णाविति जातिपरो निर्देशः । दीर्घग्रहणं मास्तु इति तदाशयः । शामिता । शमिता । शमयिता । शामिष्यते । शमिष्यते । शमयिष्यते । यङन्ताण्णिच् । शंशम्यते । शंशामिता । शंशमिता । शंशमयिता । यङ्लुगन्ताण्णिच्यप्येवम् । भाष्यमते तु यङन्ताच्चिण्वदिटि दीर्घो नास्तीति विशेषः । ण्यन्तात्वाभावे शस्यते मुनिना ॥

Rule. Take a ṇi that is followed by ciṇ or by ṇamul. Before it, the penultimate vowel of a mit root is optionally long. Ṇamul is a kṛt affix, taught later.

The sūtra says only "before ciṇ or ṇamul, a long vowel, optionally". The rest is carried forward: "before ṇi" from SK 2604, "of the penultimate" from SK 2364, and "of mit roots" from SK 2568. One of Bhaṭṭoji's commentators gives these.

Example.

śam + ṇic → vṛddhi, then shortening because śam is mit (SK 2568) → śami

→ luṭ, in the karma → śami + tās + …

→ SK 2757: treated as before ciṇ, with the joint iṭ. So the ṇi is followed by ciṇ

→ this rule: the penultimate a is optionally long → śāmi

→ the ṇi is deleted before the joint iṭ, as in bhāvitā → śāmitā.

In the other option the a stays short → śamitā.

Without the treatment of SK 2757 → śamayitā.

Likewise śāmiṣyate / śamiṣyate / śamayiṣyate.

Applies when / blocked by. Why does the sūtra say "long"? Could it not simply make the short vowel of SK 2568 optional? Bhaṭṭoji answers no. That would fail where the ṇi follows a stem that already ends in ṇi. There the first ṇi is deleted. In a rule of shortening, the deleted ṇi counts as still present (sthānivat), and it stands between. In a rule of lengthening it does not count, so the lengthening goes through. The Bālamanoramā names SK 50 for sthānivat, and SK 51 (na padānta…), which bars it in a rule of lengthening. The Tattvabodhinī also points to SK 51.

Bhaṭṭoji then reports the Bhāṣya. The Bhāṣya relies on a principle: a substitute is not sthānivat in a rule of the section that is asiddha to what precedes it (pūrvatrāsiddhīya). With that principle it rejects several words of SK 51 as superfluous: doubling, savarṇa, anusvāra, lengthening, jaś and car. On that view "ṇau" names the class "ṇi-ness" (jātipara), present in both *ṇi*s. So the Bhāṣya would drop a mention of "long". Which mention, the one in SK 51 or the one in this sūtra, scholars discuss this.

Yaṅ, then ṇic. From the intensive śaṃśamya + ṇic: śaṃśamyate; śaṃśāmitā / śaṃśamitā / śaṃśamayitā. From the yaṅluk stem + ṇic, the same.

On the Bhāṣya's view there is one difference. From the yaṅ-stem, with the joint iṭ, there is no lengthening. Bhaṭṭoji's commentators both explain: the lost a of yaṅ counts as still present and keeps the ṇi from following directly. One of them concludes that, since the results differ, the Bhāṣya's rejection is not right; scholars discuss this.

Without ṇic. Śam alone takes no object, so its lakāra is in bhāva: śamyate muninā "there is calm on the sage's part". Bhaṭṭoji's commentators both give this reason.

Watch out. The rule needs ṇi. For plain śam before ciṇ, see SK 2763.

SK 2763 · नोदात्तोपदेशस्य मान्तस्यानाचमेः · nodāttopadeśasya māntasyānācameḥ (A 7.3.34)

नोदात्तोपदेशस्य मान्तस्यानाचमेः

nodāttopadeśasya māntasyānācameḥ · A 7.3.34

उपाधाया वृद्धिर्न स्याच्चिणि ञिति णिति कृति च । अशमि । अदमि । उदात्तोपदेशस्येति किम् । अगामि । मान्तस्य किम् । अवादि । अनाचमेः किम् । आचामि ।

अनाचमिकमिवमीनामिति वक्तव्यम्

(वार्तिकम्) ॥ चिणि आयादयः -SK 2305 इति णिङभावे । अकामि । णिङ्णिचोरप्येवम् । अवामि । वध हिंसायाम् । हलन्तः । जनिवध्योः SK 2512 इति न वृद्धिः । अवधि । जाग्रोऽविचिण्णल्ङित्सु SK 2480 इत्युक्तेर्न गुणः । अजागारि ॥

Rule. Take a root that ends in m and is udātta in upadeśa, other than ācam. Its penultimate vowel takes no vṛddhi before ciṇ, or before a kṛt affix marked with ñ or ṇ.

Carried forward: "vṛddhi" from SK 2473, "of the penultimate" from SK 2282, "before ciṇ or a kṛt" from SK 2761, and "marked with ñ or ṇ" from SK 254, for the kṛt only. One of Bhaṭṭoji's commentators gives these. He adds that "udātta in upadeśa" means every root not listed as anudātta, and that ācam is cam with ā.

Example.

śam + luṅ, in bhāva → a + śam + ciṇ + ta

→ SK 2282 would give vṛddhi of the penultimate a before the ṇ-marked ciṇ

→ this rule: śam ends in m and is udātta in upadeśa, so no vṛddhi

→ the ta is deleted (SK 2329) → aśami.

Likewise dam: adami.

Applies when / blocked by. Bhaṭṭoji's counter-examples:

Why "udātta in upadeśa"? agāmi: gam is anudātta.
Why "ending in m"? avādi: vad ends in d.
Why "except ācam"? ācāmi.

A vārttika widens the exception: "It should be stated: except ācam, kam and vam." So kam and vam take vṛddhi too.

Kam normally takes ṇiṅ, but before an ārdhadhātuka it is optional (SK 2305). Without ṇiṅ: akāmi. With ṇiṅ or ṇic, the ṇi is deleted and the form is the same. Bhaṭṭoji's commentators both explain this.
Vam: avāmi.

Two more forms with ciṇ, which Bhaṭṭoji adds:

The root vadh "to kill" (listed as vadha) ends in a consonant. SK 2512, janivadhyoś ca, forbids its vṛddhi: avadhi. One of Bhaṭṭoji's commentators says this is a root of its own, not the vadha that replaces han (SK 2761).
Jāgṛ: SK 2480 excludes ciṇ, so the root takes no guṇa from that rule. Its vowel then takes vṛddhi (SK 254): ajāgāri.

Watch out. The prohibition is for the penultimate vowel only. Śam in SK 2762 has its own rule for length before ṇi.

F. The root's nasal before ciṇ: n-loss for bhañj, num for labh — running example: `bhañj + luṅ ciṇ → (n elided, optional) abhāji / abhañji`; `labh + ciṇ → (num, optional) alambhi / alābhi`

Two roots change their nasal before ciṇ. Bhañj "to break" may lose its n: abhāji, or abhañji. Labh "to obtain" may gain one, the augment num: alambhi, or alābhi. After an upasarga, labh must take it: prālambhi.

SK 2764 · भञ्जेश्च चिणि · bhañjeśca ciṇi (A 6.4.33)

भञ्जेश्च चिणि

bhañjeśca ciṇi · A 6.4.33

नलोपो वा स्यात् । अभाजि । अभञ्जि ॥

Rule. For bhañj too, when ciṇ follows, the n is optionally elided. Bhaṭṭoji's gloss: "let elision of n optionally occur".

"N" and "elision" are carried forward from SK 2544, śnān nalopaḥ. "Optionally" comes from SK 3330, taught later. One of Bhaṭṭoji's commentators gives this.

Example.

bhañj + luṅ, in the karma → a + bhañj + ciṇ + ta

→ this rule: the n (heard as ñ) is elided → a + bhaj + i + ta

→ now a is the penultimate, and takes vṛddhi before the ṇ-marked ciṇ (SK 2282) → a + bhāj + i + ta

→ the ta is deleted (SK 2329) → abhāji.

In the other option the n stays → abhañji.

One of Bhaṭṭoji's commentators states the vṛddhi step.

Watch out. The vṛddhi depends on the elision. With the n kept, a is not the penultimate, so abhañji has a short a.

SK 2765 · विभाषा चिण्णमुलोः · vibhāṣā ciṇṇamuloḥ (A 7.1.69)

विभाषा चिण्णमुलोः

vibhāṣā ciṇṇamuloḥ · A 7.1.69

लभेर्नुमागमो वा स्यात् । अलम्भि । अलाभि । व्यवस्थितविकल्पत्वात्प्रादेर्नित्यं नुम् । प्रालम्भि । द्विकर्मकाणां तु । गौणे कर्मणि दुह्यादेः प्रधाने नीहृकृष्वहाम् ॥ बुद्धिभक्षार्थयोः शब्दकर्मकाणां निजेच्छया ॥ प्रयोज्यकर्मण्यन्येषां ण्यन्तानां लादयो मताः । गौर्दुह्यते पयः । अजा ग्रामं नीयते । ह्रियते । कृष्यते । उह्यते । बोध्यते माणवकं धर्मः । माणवके धर्ममिति वा । देवदत्तो ग्रामं गम्यते । अकर्मकाणां कालादिकर्मकाणां कर्मणि भावे च लकार इष्यते । मासो मासं वा आस्यते देवदत्तेन । णिजन्तात्तु प्रयोज्ये प्रत्ययः । मासमास्यते माणवकः ॥

Rule. Labh optionally takes the augment num when ciṇ or ṇamul follows. Num is heard as n, placed after the root's vowel. "Of labh" is carried forward from SK 2582, and "num" from SK 2262. Both of Bhaṭṭoji's commentators give "of labh"; the Bālamanoramā also gives num from SK 2262.

Example.

labh + luṅ, in the karma → a + labh + ciṇ + ta

→ this rule: num → a + lambh + i + ta; the n of num is heard as m before bh

→ the a is no longer the penultimate, so it takes no vṛddhi, as in abhañji

→ the ta is deleted (SK 2329) → alambhi.

In the other option there is no num. The penultimate a takes vṛddhi → alābhi.

Applies when / blocked by. The option is fixed in its domain (vyavasthitavikalpa). After pra and the rest, num is obligatory: prālambhi. One of Bhaṭṭoji's commentators takes "pra and the rest" as all the upasargas, and traces this to a vārttika and the Bhāṣya.

With ṇamul, one of Bhaṭṭoji's commentators gives lābhaṃlābham without num. After pra only pralambhaṃpralambham.

Watch out. An option fixed in its domain is not a free choice. Without an upasarga both forms are right; with one, only the num-form.

G. 2765 (continued): which karma the lakāra goes to with dvikarmaka roots, and with time-objects — running example: `duh (gauṇa karma gauḥ expressed) → gaur duhyate payaḥ`

Some roots take two objects (dvikarmaka). "He milks the cow of milk" has gām and payaḥ. When the lakāra is in the karma, which of the two does the ending speak of? Bhaṭṭoji answers in a verse and a half, under SK 2765. Then he adds a statement on roots that take no object but have a time-word as karma. The running example is duh "to milk": gaur duhyate payaḥ. The ending speaks of the cow.

Under SK 2765: the verse on roots with two objects

The verse. Bhaṭṭoji gives it in a verse and a half:

"In the secondary (gauṇa) karma for duh and the rest; in the principal (pradhāna) karma for nī, hṛ, kṛṣ and vah. For roots meaning cognition (buddhi) or eating (bhakṣa), and for roots whose karma is speech, at one's own choice. In the karma that is the one caused to act (prayojya), for the others, when they end in ṇi. There the lakāras and the rest are held to come."

Bhaṭṭoji's commentators both explain its terms.

The secondary karma is the one named karma by SK 539, akathitaṃ ca. The principal karma is named by another rule.
"Duh and the rest" are twelve roots: duh, yāc, pac, daṇḍ, rudh, pracch, ci, brū, śās, ji, math, muṣ.
For the roots of SK 540, the secondary karma is the one that sūtra names.
"The others" are roots of motion, roots without an object, and hṛ and kṛ of SK 541.
"The lakāras and the rest" also covers the kṛtya affixes, kta and the khalartha affixes.

One of them, the Bālamanoramā, adds that the verse sums up the vārttikas and the Bhāṣya on SK 539.

Example.

duh: the milk (payaḥ) is the principal karma; the cow (go) is karma by SK 539, the secondary one

→ the verse: for duh, the lakāra goes to the secondary karma

→ the ending expresses the cow, so it takes the first case: gauḥ

→ the milk is not expressed, so it keeps the second case: payaḥ

→ gaur duhyate payaḥ "the cow is milked of [her] milk".

Bhaṭṭoji's other examples:

The principal karma, for nī: ajā grāmaṃ nīyate "the goat is led to the village". Likewise hriyate, kṛṣyate, uhyate.
At choice, for cognition: bodhyate māṇavakaṃ dharmaḥ "dharma is taught to the boy", or māṇavako dharmam "the boy is taught dharma".
The one caused to act, for a causative of motion: devadatto grāmaṃ gamyate "Devadatta is made to go to the village".

Time as karma. Bhaṭṭoji adds: for roots without an object that have time and the like as karma, the lakāra is wanted in the karma and in bhāva. So māso māsaṃ vā āsyate devadattena: "the month is sat through by Devadatta", or "there is sitting for a month by Devadatta".

One of Bhaṭṭoji's commentators notes that this statement is not a vārttika; it rests on reasoning. He also reads the bhāva form with the seventh case, māse; scholars discuss this.

From the causative, the affix goes to the one caused to act: māsam āsyate māṇavakaḥ "the boy is made to sit for a month". One of Bhaṭṭoji's commentators says the prompter, devadattena, is to be supplied.

So ends the section on bhāva and karma.

Watch out. With two objects, only one is expressed by the ending and stands in the first case. The other keeps the second case, as in gaur duhyate payaḥ.

Terms used in this chapter

karman (karma) — the object; a lakāra "in the karma" speaks of the object (SK 2152).
bhāva — the bare action itself; a lakāra "in bhāva" speaks of the action. Bhaṭṭoji glosses it bhāvanā, "bringing about".
lakāra — one of the ten abstract affixes of tense and mood, all reduced to l.
taṅ / ātmanepada — the ātmanepada endings, which every lakāra in bhāva or karma takes (SK 2679).
yak — the affix ya, marked with k, between the root and a sārvadhātuka ending in bhāva or karma (SK 2756).
dhātu — a root.
sārvadhātuka / ārdhadhātuka — the name of the verb-endings and ś-marked affixes (SK 2166) / of the other affixes after a root (SK 2187).
kit — marked with k; such an affix blocks guṇa and vṛddhi of the root vowel.
anūdyate — "is restated": said of a lakāra in bhāva, which repeats the action the root already denotes.
yuṣmad / asmad — the pronoun stems "you" / "I".
sāmānādhikaraṇya — co-reference: two words naming the same thing.
asattva — not a substance; such a meaning has no number of its own.
autsargika — general, the default; said of the singular ending.
anabhihita — not expressed by the ending; such an agent takes the third case.
anuvṛtti — carrying a word forward from another sūtra.
ciṇ — the replacement of cli, heard as i, before ta in bhāva or karma (SK 2758); the ta after it is deleted (SK 2329).
cli / sic — the abstract aorist affix / its usual replacement, heard as s (SK 2222).
aṭ — the augment a put before the aṅga in luṅ, laṅ and lṛṅ (chapter 44): a + bhū.
sya / tās / sīyuṭ — the affix of lṛṭ and lṛṅ / the affix of luṭ / the augment of the ātmanepada liṅ endings.
ciṇvat, ciṇvadbhāva — "as before ciṇ": the optional treatment of SK 2757, under which an aṅga takes the operations it would take before ciṇ.
aṅga / aṅgakārya — the base to which an affix is added / an operation on that base; only these are brought by the treatment.
upadeśa — the form as first taught; SK 2757 needs "a vowel that stands in upadeśa".
iṭ — the augment i. The joint iṭ of SK 2757 comes with the treatment; the ordinary iṭ comes before a val-initial ārdhadhātuka (SK 2184).
val — any consonant except y; a val-initial affix begins with such a consonant.
seṭ — a root that takes the augment iṭ.
sanniyogaśiṣṭa — enjoined jointly: present only together with another operation.
vṛddhi / guṇa — ā, ai, au / a, e, o.
vidhiliṅ / āśīrliṅ — the liṅ of injunction, whose endings are sārvadhātuka / the liṅ of blessing, whose endings are ārdhadhātuka.
viṣayasaptamī — a seventh case of domain: "where bhāva or karma is understood".
nitya — a rule that applies whatever is done first; it wins.
ṇi / ṇic / ṇyanta — the affix i of a causative or curādi stem / the causative affix / a stem ending in ṇi.
ābhīya — a section of aṅga rules whose results count as asiddha among themselves. The joint iṭ belongs to it, so SK 2313 does not see it.
asiddha — treated as not having taken effect.
san / yaṅ / yaṅluk — the desiderative affix / the intensive affix / the intensive stem whose yaṅ is lost.
rapara — with r added after: said of guṇa ar for ṛ.
para — later in the order of rules; it normally wins.
idit — a root marked with i; it keeps its n (SK 415 does not act).
saṃprasāraṇa — the change of a semivowel to its vowel, as in yaj → ij.
anutāpa — remorse (SK 2760).
karmakartṛ — a karma that is itself the agent; taught in chapter 64.
karmadhāraya — a compound whose members name the same thing; aśiti read so means "before an affix whose initial is a marker ś".
eś — the substitute e for ta in liṭ (SK 2241); its ś is a final marker.
upasarga — a prefix such as anu, pra, construed with a verb.
yuk — the augment y for an ā-final aṅga before ciṇ or a ñ/ṇ-marked kṛt (SK 2761).
ñit / ṇit — marked with ñ / with ṇ.
kṛt — an affix added to a root to make a noun or participle.
taddhita — an affix added to a noun, such as iñ.
apavāda — an exception, which sets aside a more general rule.
niravakāśa — "without other scope": a rule clause that would do nothing unless it overrides.
vadha — the substitute for han before liṅ (SK 2433).
latva / ḍhatva — l for r / ḍh for dh.
hetumaṇṇic — Bhaṭṭoji's name for the causative ṇic.
upadhā — the penultimate sound.
mit — a root whose vowel stays short before ṇi (SK 2568).
ṇamul — a kṛt affix, taught later, before which SK 2762 and SK 2765 also act.
sthānivat — treated like the original it replaced; a deleted ṇi may still count as present.
pūrvatrāsiddhīya — belonging to the section that is asiddha to what precedes; the Bhāṣya bars sthānivat in such rules.
savarṇa / anusvāra / jaś / car — names of four sandhi operations listed in SK 51: a substitute of the same kind, the after-vowel nasal ṃ, a voiced unaspirated stop, and a voiceless unaspirated stop or sibilant.
jātipara — referring to a class, not one item.
Bhāṣya — Patañjali's great commentary.
udātta / anudātta — high / low accent; a root's accent in upadeśa decides SK 2763.
vārttika — a supplementary statement that corrects or completes a sūtra.
ṇiṅ — the affix i marked with ṅ, taken by kam (optionally before an ārdhadhātuka, SK 2305).
num — the augment n placed after a root's vowel.
vyavasthitavikalpa — an option fixed in its domain: optional in one setting, fixed in another.
dvikarmaka — a root with two objects.
gauṇa / pradhāna karma — the secondary object (named by SK 539) / the principal object.
prayojya — the one caused to act, in a causative.
kṛtya / kta / khalartha — kinds of kṛt affix that, like the lakāras, follow the verse in SK 2765.
Bālamanoramā / Tattvabodhinī — the two commentaries on Bhaṭṭoji's Kaumudī that this chapter calls "Bhaṭṭoji's commentators".

Summary

SK	What it does	Example
2756	yak before a sārvadhātuka in bhāva or karma; bhāva is third singular only, agent in the third case	`tvayā bhūyate`; liṭ `babhūve`
2757	before sya, sic, sīyuṭ, tās in bhāva/karma, the aṅga of a vowel-root (vowel in upadeśa), han, grah, dṛś is optionally treated as before ciṇ, with a joint iṭ	`bhāvitā / bhavitā`, `bhāviṣyate / bhaviṣyate`
2758	ciṇ replaces cli before ta in bhāva/karma; survey of yak forms (karma, causative, desiderative, intensive, stem changes)	`abhāvi`; `anubhūyate`; `bhāvyate`, `bhāvitā / bhāvayitā`; `stūyate`, `āritā / artā`
2759	optional ā for the final of tan before yak	`tāyate / tanyate`; `jāyate / janyate`
2760	no ciṇ for tap when the karma is agent, and in remorse; also `dīyate`, `jagle`	`anvatapta pāpena`
2761	augment yuk for an ā-final aṅga before ciṇ or a ñ/ṇ-marked kṛt; forms of han, grah, dṛś; the 104 forms of gṝ	`dāyitā / dātā`, `adāyi`; `ghānitā / hantā`; `grāhitā / grahītā`; `adarśi`; `śamyate`
2762	optional long penultimate of a mit root before ṇi + ciṇ/ṇamul	`śāmitā / śamitā / śamayitā`
2763	no penultimate vṛddhi before ciṇ/ñ-ṇ kṛt for an m-final udātta root, except ācam (vārttika: and kam, vam)	`aśami`, `adami`; `ācāmi`, `akāmi`, `avāmi`
2764	optional loss of the n of bhañj before ciṇ	`abhāji / abhañji`
2765	optional num for labh before ciṇ/ṇamul, obligatory after an upasarga; then the verse on two-object roots, and time as karma	`alambhi / alābhi`, `prālambhi`; `gaur duhyate payaḥ`; `māsam āsyate māṇavakaḥ`

← textbook contents reading page (Technical) →

63. Bhāvakarma-tiṅ-prakaraṇa — Verbs in bhāva and karma: yak, ciṇ and the ciṇvat-treatment

Before you start

A. yak in bhāva and karma: the impersonal third singular — running example: bhū + laṭ (bhāva) → ta (taṅ, SK 2679) → bhū + yak + te → bhūyate; tvayā mayā 'nyaiś ca bhūyate

SK 2756 · सार्वधातुके यक् · sārvadhātuke yak (A 3.1.67)

B. ciṇ in the luṅ and the ciṇvat-treatment, with a survey of yak forms — running example: bhū + luṅ (bhāva), ta → cli → ciṇ → (vṛddhi, ta luk) → abhāvi; bhū + tās (ciṇvat: vṛddhi + iṭ) → bhāvitā / bhavitā

SK 2758 · चिण् भावकर्मणोः · ciṇ bhāvakarmaṇoḥ (A 3.1.66)

C. Two root-specific adjustments, to yak and to ciṇ — running example: tan + yak → (ā for n, optional) → tāyate / tanyate; anu + tap, luṅ → (no ciṇ) sic → anvatapta pāpena

SK 2759 · तनोतेर्यकि · tanoteryaki (A 6.4.44)

SK 2760 · तपोऽनुतापे च · tapo'nutāpe ca (A 3.1.65)

D. yuk for ā-finals, and the ciṇvat forms of han, grah, dṛś and gṝ — running example: dā + tās (ciṇvat) → dā-y-i-tā (yuk) → dāyitā / dātā; dā + luṅ ciṇ → adāyi

SK 2761 · आतो युक् चिण्कृतोः · āto yuk ciṇkṛtoḥ (A 7.3.33)

E. The upadhā vowel before ciṇ: optional length for mit roots, no vṛddhi for m-final udātta roots — running example: śam + ṇic + tās (ciṇvat) → śāmitā / śamitā / śamayitā; śam + luṅ ciṇ → (no vṛddhi) aśami

SK 2762 · चिण्णमुलोर्दीर्घोऽन्यतरस्याम् · ciṇṇamulordīrgho'nyatarasyām (A 6.4.93)

SK 2763 · नोदात्तोपदेशस्य मान्तस्यानाचमेः · nodāttopadeśasya māntasyānācameḥ (A 7.3.34)

F. The root's nasal before ciṇ: n-loss for bhañj, num for labh — running example: bhañj + luṅ ciṇ → (n elided, optional) abhāji / abhañji; labh + ciṇ → (num, optional) alambhi / alābhi

SK 2764 · भञ्जेश्च चिणि · bhañjeśca ciṇi (A 6.4.33)

SK 2765 · विभाषा चिण्णमुलोः · vibhāṣā ciṇṇamuloḥ (A 7.1.69)

G. 2765 (continued): which karma the lakāra goes to with dvikarmaka roots, and with time-objects — running example: duh (gauṇa karma gauḥ expressed) → gaur duhyate payaḥ

Under SK 2765: the verse on roots with two objects

Terms used in this chapter

Summary

A. yak in bhāva and karma: the impersonal third singular — running example: `bhū + laṭ (bhāva) → ta (taṅ, SK 2679) → bhū + yak + te → bhūyate`; `tvayā mayā 'nyaiś ca bhūyate`

B. ciṇ in the luṅ and the ciṇvat-treatment, with a survey of yak forms — running example: `bhū + luṅ (bhāva), ta → cli → ciṇ → (vṛddhi, ta luk) → abhāvi`; `bhū + tās (ciṇvat: vṛddhi + iṭ) → bhāvitā / bhavitā`

C. Two root-specific adjustments, to yak and to ciṇ — running example: `tan + yak → (ā for n, optional) → tāyate / tanyate`; `anu + tap, luṅ → (no ciṇ) sic → anvatapta pāpena`

D. yuk for ā-finals, and the ciṇvat forms of han, grah, dṛś and gṝ — running example: `dā + tās (ciṇvat) → dā-y-i-tā (yuk) → dāyitā / dātā`; `dā + luṅ ciṇ → adāyi`

E. The upadhā vowel before ciṇ: optional length for mit roots, no vṛddhi for m-final udātta roots — running example: `śam + ṇic + tās (ciṇvat) → śāmitā / śamitā / śamayitā`; `śam + luṅ ciṇ → (no vṛddhi) aśami`

F. The root's nasal before ciṇ: n-loss for bhañj, num for labh — running example: `bhañj + luṅ ciṇ → (n elided, optional) abhāji / abhañji`; `labh + ciṇ → (num, optional) alambhi / alābhi`

G. 2765 (continued): which karma the lakāra goes to with dvikarmaka roots, and with time-objects — running example: `duh (gauṇa karma gauḥ expressed) → gaur duhyate payaḥ`